W3C home > Mailing lists > Public > w3c-sgml-wg@w3.org > February 1997

Re: 1.a: Use Elements? -- critical ambiguity in question!

From: Eve Maler <eve@doctools.com>
Date: Sun, 02 Feb 1997 21:46:37 -0500
Message-Id: <199702030246.VAA08323@village.doctools.com>
To: w3c-sgml-wg@www10.w3.org
>Eve, your idea is an interesting one:
>> <?XML ATTLIST PrimaryIE
>>         xml             (linkto)        "linkto"
>>         xmlnames        CDATA           #FIXED "linkends ptr"
>>         scheme          (intid)         "intid"
>> ?>
>
>but I think I see a problem with it -- XML documents would use these
>extra attributes, but they would not be declared to the SGML parser,
>and the result would not be valid SGML.
>
>Now, at one point, a partial DTD wasn't allowed -- it had to be all
>or nothing.  If that's still the case, the rules on undeclared entities 
>are inconsistent.  But if it isn't the case, I think we must either
>(1) apply pressure for multiple Attlists, and help draw up a draft
>    amendment if necessary;
>
>(2) until then, stick to markup that is clearly different.
>
>    <?-XML- URI element A attribute HREF?>
>    <?-XML- URI element FOOTNOTE attribute HREF?>
>    <?-XML- SYSTEM element IMAGE attribute SRC action embed?>
>    or even
>    <?-XML-XHL- -XML-XHL-URI- -XML-XHL-ELEMENT A ..... ugh?>
>
>Lee

I see your point.  Does this mean that without multiple ATTLISTs,
there's no point in our even defining architectures?

Then again, if we only allowed #FIXED attributes (and possible the
other "fixed" token value method that I used above) in the ATTLIST
PI, and specified that these attributes not appear in the instance,
SGML wouldn't have a problem with it.  Yuck...too many if's.  I
think I just talked myself out of it.

	Eve
Received on Sunday, 2 February 1997 21:47:03 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:25:07 UTC